The concept of naming conventions is paramount to the creation of "enterprise" networking solutions and will effect your entire infrastructure concerning maintenance, anomaly resoluition, performance, security, training, regulatory compliance and scalability. I think that covers more than most expect hear. Remember this is just a short answer. Let's talk about UNIX based environments in support of medium to large scale healthcare enterprise networking solution (Infranet). I anticipate more questions to follow, so lets begin this from the beginning. UNIX operating systems are designed fortunately and unfortuntely to be highly configurable. The understanding of how to engineer user "accounts", "groups" and "permissions" will make or break your engineering solution, sometimes sooner or sometimes later. If you are lucky, it will be sooner than later. Sooner is good because it will break before you go live and be more embarrassing than expensive. Later, if you only use vendor defaults, because it will always mostly work, but will be much like that of your college campus, not regulatory compliant, hard to map enterprise monitoring solutions, impossible to add OLAP solutions, hard to train operations personnel, forget about security, and oh? you forgot to design an efficient software development life cycle. Conclusion, not designing a good naming convention will hide all your inefficiencies and people will leave because no one will be able to fix it (most who will leave will leave on purpose). Designing a proper naming convention will remove the drama, almost make it boring, and all the promises of using advanced computer technologies will be easier and less expensive to maintain, engineer and plan for. (100,000 foot answer - jzr).